Research On The Energy-saving Technology Of Linecard Of Mixed Optical/electric Switching Equipments

The number of electrical switching units is less than that of optical switching units in the optical/electrical maxed switching devices. It can be explained by two reasons: one is high power consumption of electrical switching units; another one is that the complexity and the cost of large-scale multi-stage switching network composed of electrical switching units are large. So the number of electrical switching units should be shortened. However, lots of wavelengths require more electrical port to groom. If electrical ports are used sparingly when the blocking probability is ensured, it can save energy. This paper researches on port allocation under low traffic load to reduce power consumption.The problem of linecard port assignment is not considered before, because of its simple that we only need to find and assign free linecard port for a new lightpath. There are two traditional algorithms, FF and MUF, which both have low linecard utilization rate although FF is better than MUF due to the randomness of FF. based on FF and MUF, a compression algorithm of linecard is proposed. Simulation results show that the compression algorithm can effectively improve the utilization of linecard despite it is simple. To get better performance of the linecard, three algorithms are designed. One is the MERGE algorithm based on the same established lightpath. The second one is COOPER_band algorithm, the thrid is COOPER_port algorithm. Because of no change in the paths of traffic, MERGE is simple but the effect is obvious. However, in the last two algorithms the paths of traffic changed by re-routing, which is a random process, so that the entire network traffic alters. Simulation results have shown that the last two algorithms are further improved from the traditional algorithms, and can get a good result of saving power up to50%.Comparing with high traffic load, the power consumption of the entire network under small traffic load has litter difference, probably at the ratio of1:2.By classifying all the nodes into the primary nodes and the secondary node, only the linecards in the primary nodes be turned on and the secondary'turned off. On the condition that the blocking probability is zero case, it is actually a compression strategy. Eventually, we can further reduce the power consumption under low traffic load so as to the ratio of low traffic load to high traffic is1:4. In addition, the algorithm has up to60%improvement on energy-saving over conventional algorithms.